DDLS Course 2025: Computer Lab 1 Feedback Summary

Last updated on Sep 3, 2025 4 min read

Thanks to everyone for the detailed feedback and peer reviews. Below is a summary of what worked well, where students struggled, and a consolidated FAQ with answers.

What Worked Well

Clarity & structure. Many students found the manual easy to follow, with clear step-by-step instructions.
Hands-on usefulness. Exposure to AI tutors, CLI agents, and workflows felt immediately relevant to research.
Organization discipline. Submissions with clear headers, comments, and well-structured files were praised.
Engagement. Several students explored beyond the tutorial—trying new optimizers, testing clustering methods, or building bonus CNNs.

“Everything is amazingly well described… I feel like my basic understanding is very rigid now.” “The computer lab was great; I’ll keep exploring at home.” “AI tutor is a fun concept—and useful.”

Where Students Struggled

Access & stability issues. Gemini CLI login, Colab disconnects, slow VS Code↔Colab connections, and model switches caused frustration.
Deliverables. Some submissions were missing chat logs, executed notebooks, or exported figures. Multiple chat logs also caused confusion.
Process. A few students felt the tutorial was too long or answered its own questions, leading to more copy-pasting than reflection.
Expectations. Beginners worried about fairness in grading compared to experienced coders.
Orientation. Requests for a short VS Code/CLI primer and clearer instructions on pre-lab preparation.

Our Approach to Questions

Many of the most frequent questions are technical setup issues. These are important learning opportunities:

You should first try to solve them yourself—ask your AI tutor, search online, or make a reasonable assumption.
Use the computer lab time to get help if you remain stuck.
Remember: our goal is not to babysit every detail, but to train your problem-solving skills.

For course requirement–related questions (grading, deliverables, expectations), we are happy to provide assistance and clarifications.

Student FAQ (with Answers)

Setup & Access

Gemini CLI login problems / switching accounts? Save your files in Google Drive and download/re-upload them if you switch account.
Prevent Colab disconnects / losing chat history? Save snapshots frequently.
VS Code ↔ Colab too slow? Try starting a local runtime.
Run the lab locally in VS Code? Yes, check the same local runtime guide.

AI Tutor & Workflow

AI tutor outputs too much or switches model? Tell it to shorten responses, and add this preference in your GEMINI.md file.
Tutorial answers its own questions? Add an instruction in the GEMINI.md file asking it not to.
Too much copy-pasting—can we do “fill-in-the-gaps”? Ask the tutor to design tasks that way and note it in your GEMINI.md.

Deliverables & Organization

What are the required deliverables? Notebook, chat log, README, outputs. Baseline is to describe the folder structure in the README.
Missing a file—what’s the minimal checklist? You should submit all the files mentioned in the computer lab instructions.
Multiple chat logs—how to pick final? Indicate in the README, or remove extras.
README vs. in-notebook markdown? README = describe the folder + notes for reviewers. Notebook markdown = explain code and results.
How much commentary in the notebook? Enough to explain what you’re doing, no need to over-explain.
Exporting plots/data to outputs? Ask your AI assistant to generate the save code (e.g., plt.savefig() or np.savetxt()).

Learning & Grading

How to use AI efficiently? Watch the video on context engineering, and start with a good GEMINI.md.
I’m new to Python—will grading be fair? Yes. Labs are pass/fail, and grading focuses on what you learn, not your starting level.
Pre-lab video wasn’t clear—what to watch before? We sent an email: video 1 is pre-lab, video 2 can be followed during the lab.
VS Code/CLI primer to reduce setup confusion? Covered in the video recording—please watch it.
Not comfortable sharing chat history? Sharing helps peers give feedback on prompt skills. But from Lab 2, you can opt out of peer review if you don’t want to share.

Course Context

Couldn’t open notebook/figures in Colab? This can happen—try restarting runtime or saving figures manually.
How do outputs connect to real-world applications? We’ll discuss this in later modules (bioinformatics, drug discovery, etc.).

👉 Remember: The main goal of the computer lab is not just to follow instructions but to practice problem-solving. Many details are intentionally left open so you can build confidence in exploring, debugging, and asking the right questions.